diabetic ketoacidosisInsulin does MANY things in the body, but the role we care about in the Emergency Department is glucose regulation. Insulin allows cells to take up glucose from the blood stream, inhibits liver glucose production, increases glycogen storage, and increases lipid production. When insulin is not present, such as in patients with Type 1 diabetes mellitus (DM), all of the opposite effects occur.


A lack of insulin causes the following downstream effects:

  • Prevents glucose from being used as an energy source – Free fatty acids are used instead and produce ketoacids during metabolism.
  • Causes a surge of stress hormones and induces gluconeogenesis – When blood glucose levels are elevated, the kidneys cannot absorb all of the glucose from the urine, and the extra glucose in the urine causes polyuria, even in the setting of dehydration. In addition, acidosis causes potassium to shift out of cells into the blood, and the combination of this with dehydration causes the body to preferentially retain sodium at the expense of potassium.1,2

When insulin homeostasis is disrupted and decompensates, patients are at risk for developing diabetic ketoacidosis (DKA).

Definition of DKA1

All of the following criteria are required for a diagnosis of DKA:

  1. Hyperglycemia (glucose >200 mg/dL)
  2. Acidosis (pH <7.3 or bicarb <15 mmol/L)
  3. Ketosis (by urine or blood test)

Treatment of DKA

Treatment is based on a simple principle: return the body’s glucose regulation to its normal state and replace all of the things the body consumed while insulin-deficient. While bolus insulin is common in the treatment of DKA in adults, it is relatively contraindicated in the pediatric patient. Dehydration and secondary sympathetic activation can interfere with local tissue perfusion and may cause irregular and unpredictable absorption.

Step 1: Correction of shock

  • If the patient has hypovolemic shock, crystalloid fluid boluses are indicated. However, many experts recommend expanding volume gradually to reduce the risk of cerebral edema.3,4

Step 2: Fluids, glucose, and insulin using the 2-Bag Method

FLUIDS: Set total fluid rate at 1.5 X maintenance intravenous fluid rate (MIVF). The fluids come from 2 separate bags (saline bag, glucose bag).

    • Bag #1 (saline bag): Crystalloid
      • Institutions vary in the use of crystalloid formulations (NS, ¾ NS, or ½ NS).
      • Because the whole body is expected to be potassium depleted, 40 mEq/L of potassium should be added to the fluids*. A common combination is to use 20 mEq of K-Phosphate and 20 mEq of K-Acetate. Because KCl is an additional chloride load that can cause metabolic acidosis, it should be avoided or limited when possible.
    • Bag #2 (glucose bag): D10
      • Other than glucose, it should have the same electrolyte composition as Bag #1.

*NOTE on potassium: In severe DKA, the potassium can be profoundly low, and starting insulin before repleting potassium may precipitate symptomatic hypokalemia. Some experts recommend not starting insulin until the potassium has been corrected to above 3.0 mmol/L. If the initial measured potassium is elevated, or the patient is not urinating, no potassium should be added to the fluids until the level is <5.5 mmol/L and the patient is urinating.1,2

INSULIN: Once these fluids are running at 1.5 x MIVF, an insulin infusion can be started at 0.1 units/kg/hr.

INFUSION CALCULATIONS: When starting the 2-Bag Method approach, use the table below for the rate of each bag.

  • Example 1: If the initial blood glucose is >350 mg/dL, do not give dextrose. 100% of the fluid should come from the saline bag. Every hour check the glucose and adjust the fluids, based on the glucose level.
Blood Glucose
Saline Bag
(% of Total Fluid Rate)
Glucose Bag
(% of Total Fluid Rate)
Dextrose %
  • Example 2: A 20-kg child presents in DKA with a glucose of 325 mg/dL

Reasoning Behind the 2-Bag Method

As the blood glucose levels decline, the rate of D10 “glucose bag” infusion should increase to avoid hypoglycemia. Concurrently, the saline bag infusion rate is slowed to maintain the same total 1.5 x MIVF rate. A common pitfall is to reduce the insulin infusion when the patient’s glucose levels get closer to 200 mg/dL. Instead glucose should be added to the infusion to help replete the body’s intracellular glucose stores, thus halting the production of ketoacids and the improving the patient’s bicarbonate and pH.


Despite the profound acidosis present in DKA, bicarbonate should not generally be administered. Giving bicarbonate to drive pH to near a normal level will cause a significant rise in CO2, which then lowers brain pH as CO2 crosses the blood brain barrier. Bicarbonate can also cause metabolic alkalosis as the insulin infusion reduces ketoacidosis and regenerates endogenous bicarbonate. Separately, bicarbonate has been associated with development of cerebral edema and hypokalemia.3,5 Reserve treatment for those with cardiac dysfunction associated with a pH <6.9 or if there is severe hyperkalemia.

Step 3: Monitoring

  • Monitor blood glucose levels every hour
  • Alternate measuring a basic metabolic panel and blood gas (each one every other hour).
  • Monitor urine output, as potassium must be removed from the IV fluids if there is no urine output.
  • Obtain an electrocardiogram (ECG) in the setting of severe DKA or there is a concern for severe potassium derangements. Series ECGs may be warranted.

Step 4: Transitioning to subcutaneous insulin

Moving from a continuous insulin drip to subcutaneous insulin can be done when the glucose has normalized AND the acidosis has resolved. In general, this means:

  • The bicarbonate level is ≥15 mmol/L.1,2
  • The patient must be able to eat without vomiting.

Administer a dose of a subcutaneous, long-acting insulin. Within 15 minutes after administration, stop the insulin drip and check a fingerstick glucose. Continue to monitor blood glucose levels every hour until they are stable on the subcutaneous regimen.

What can go wrong?

1. Hypoglycemia6

Should the glucose begin dropping too fast, increase the percentage of dextrose containing fluids in the 2-bag system. Change the D10 to D12.5, if needed. If the glucose levels continue to drop too quickly, decrease the insulin drip. Regardless of glucose level, if a patient develops symptomatic hypoglycemia (shakiness, confusion, altered mental status), stop the insulin drip for 10 minutes and then restart either at a lower rate or with more dextrose.

2. Cerebral edema (CE)6

The pathophysiology of CE is uncertain. It is most likely to occur between 4-12 hours after initiation of therapy, but possible anytime in the first 24 hours. It is rare before therapy or after 24 hours. It occurs in 0.3–1% of children with DKA, but mortality is 21-24%, if it develops. Several theories exist regarding the development of CE.

  • Vasogenic edema: Primary damage to the cerebral vascular endothelium and increased blood brain barrier permeability or disturbance in autoregulation permit abnormal diffusion into cerebral tissues. Some MR studies show water diffusion into cerebral tissues during DKA, while others suggest abnormality in the blood brain barrier.7
  • Osmotic edema secondary to fluid therapy: IV fluids have lower osmolarity than the intracellular compartment during DKA treatment. The combination of insulin and hypotonic fluids lowers serum osmolarity, moving water into the brain via osmosis. The problem with this theory is that this phenomenon should be present in all patients with DKA, but CE is not that common. Little data exists to support this theory.8

Risk factors for CE3

  • Younger children
  • Newly diagnosis DM
  • Severity of acidosis
  • Increased BUN at presentation
  • Use of bicarbonate
  • Low pCO2 of initial sample after adjusting for severity of acidosis

Treatment of CE

  • Reduce the rate of fluid administration down to 1x MIVF.
  • Give mannitol – some studies suggest higher mortality with 3% saline (hypertonic) than mannitol in DKA-related CE. The dose for mannitol is 0.5 to 1 g/kg over 20 minutes, which can be repeated in 2 hours if no response.
  • There is no data to confirm that elevating the head of bed helps, but some practitioners do because of extrapolation from the TBI guidelines.

3. Other complications

  • Venous thrombosis
  • Aspiration (if vomiting or altered)
  • Cardiac arrhythmias (from potassium derangement)

The Bottom Line: Pediatric DKA

1. Diagnose by:

  • Signs of polyuria, polydipsia, deep fast breaths (Kussmaul)
  • Elevated glucose, low pH, low bicarbonate, ketones in blood or urine

2. Treatment

  • Bag #1: Start an IV infusion of normal saline fluids (+/- potassium).
  • Bag #2: Prepare a bag of D10 fluids (+/- potassium) comprised of the same electrolytes as Bag #1.
  • Start an IV insulin drip at 0.1 units/kg/hour.
  • Measure glucose once an hour; alternate between measuring a basic metabolic panel and blood gas every hour.
  • When the glucose drops to <350 mg/dL, infuse D10 fluids (bag #2) and slow the infusion rate of the normal saline bag (bag #1) to maintain a rate of 1.5 x MIVF (see above table for infusion rates).

3. When to transition from insulin drip to subcutaneous insulin

  • Stable and normal glucose
  • Bicarbonate ≥ 15 mmol/L
  • Patient can tolerate food

4. Complications and Troubleshooting

  • If the blood glucose is too low and the patient is symptomatic, turn off the insulin drip for 10 minutes.
  • If the glucose is still high but dropping too fast, either slow the insulin drip rate or start D12.5 crystalloid fluid.
  • If the patient becomes confused or unresponsive, worry about hypoglycemia or cerebral edema.

Updated: April 2, 2018 to address the reader comment about D10 crystalloid fluids in Bag #2.

Wolfsdorf J, Glaser N, Sperling M, American D. Diabetic ketoacidosis in infants, children, and adolescents: A consensus statement from the American Diabetes Association. Diabetes Care. 2006;29(5):1150-1159. [PubMed]
Dunger D, Sperling M, Acerini C, et al. ESPE/LWPES consensus statement on diabetic ketoacidosis in children and adolescents. Arch Dis Child. 2004;89(2):188-194. [PubMed]
Glaser N, Barnett P, McCaslin I, et al. Risk factors for cerebral edema in children with diabetic ketoacidosis. The Pediatric Emergency Medicine Collaborative Research Committee of the American Academy of Pediatrics. N Engl J Med. 2001;344(4):264-269. [PubMed]
Hsia D, Tarai S, Alimi A, Coss-Bu J, Haymond M. Fluid management in pediatric patients with DKA and rates of suspected clinical cerebral edema. Pediatr Diabetes. 2015;16(5):338-344. [PubMed]
Glaser N, Wootton-Gorges S, Marcin J, et al. Mechanism of cerebral edema in children with diabetic ketoacidosis. J Pediatr. 2004;145(2):164-171. [PubMed]
Edge J, Ford-Adams M, Dunger D. Causes of death in children with insulin dependent diabetes 1990-96. Arch Dis Child. 1999;81(4):318-323. [PubMed]
Figueroa R, Hoffman W, Momin Z, Pancholy A, Passmore G, Allison J. Study of subclinical cerebral edema in diabetic ketoacidosis by magnetic resonance imaging T2 relaxometry and apparent diffusion coefficient maps. Endocr Res. 2005;31(4):345-355. [PubMed]
Lawrence S, Cummings E, Gaboury I, Daneman D. Population-based study of incidence and risk factors for cerebral edema in pediatric diabetic ketoacidosis. J Pediatr. 2005;146(5):688-692. [PubMed]
Jason Woods, MD

Jason Woods, MD

ALiEM Podcast Editor for ACEP E-QUAL Series
Assistant Professor
Department of Pediatrics, Section of Emergency Medicine
University of Colorado, School of Medicine
Jason Woods, MD


PEM physician. Creator of @littlebigmed podcast. Views are my own and do not represent medical advice. RT≠endorsement
Josh Bukowski, MD

Josh Bukowski, MD

Captain, USAF, MC
SAMMC Department of Emergency Medicine
JBSA Fort Sam Houston
Josh Bukowski, MD


Emergency Medicine Resident at University of California San Francisco Captain United States Air Force